It was a Wednesday on my funded account, and I was short GBPUSD for the third time in a week. Price had tapped the same ceiling twice, fallen back twice, and each time I had told myself the double top was enough. The third tap came, price rejected it again, and this time I stayed out — because a friend had sent me a chart with the same level tested three times and the words “this is not a double top any more” written across it. Price broke the neckline that afternoon, ran 68 pips, and never looked back. I had missed the cleanest reversal of my month because I had stopped counting at two.
In this article
- What a triple top and a triple bottom actually are
- The three-touch test: telling a real triple from a range in disguise
- The full rules: setup, entry, stop, targets and session filter
- Position-sizing maths under a 5,000 dollar daily limit
- The worked trade: a triple top on GBPUSD
- I tested it: ten filtered triple trades against the unfiltered version
- When it fails, and how it fits prop firm rules
- Frequently asked questions
That missed trade taught me why triple tops and bottoms matter: the third touch changes the story. Two touches can be a coincidence or a range. Three failed attempts at the same level mean one side has spent its ammunition. When the neckline finally gives way, the move is often faster and cleaner than a double-top break, because three rounds of trapped traders head for the same exit. On a funded account, that speed is both the opportunity and the danger.
What a triple top and a triple bottom actually are
A triple top forms after a clear uptrend. Price rises to a resistance level, pulls back, rises to the same level again, pulls back again, and rises a third time to roughly the same level before falling away. The two pullback lows between the three peaks form the neckline. The pattern is only complete when price closes below that neckline. Until that close, you have a range with three taps on the ceiling — interesting, but not a trade.
A triple bottom is the mirror image. After a clear downtrend, price falls to a support level three times, with two relief rallies in between, and the pattern completes only when price closes above the neckline drawn through those two rally highs. Same logic, opposite direction: three failed attempts at a floor, then a break that traps three rounds of sellers.
| Feature | Triple top | Triple bottom |
|---|---|---|
| Prior trend | Clear uptrend | Clear downtrend |
| Level tested three times | Resistance ceiling | Support floor |
| Neckline | Through the two pullback lows | Through the two relief-rally highs |
| Completion | Close below the neckline | Close above the neckline |
| Direction of the trade | Short | Long |
| Measured target | Neckline minus pattern height | Neckline plus pattern height |
One distinction matters for readers of this site, because we already cover the two-touch version. The double top and double bottom strategy trades the second rejection and the neckline break that follows it. A triple pattern is a different trade, not a renamed one: it waits longer, it demands a third failure, and its neckline is drawn through two pullbacks instead of one. If you trade every double top as though it were already a triple, you will enter a third of your trades before the pattern that justifies them even exists.
The three-touch test: telling a real triple from a range in disguise
Most charts that look like triple tops are not. They are rectangles, and trading a rectangle break as a reversal is how funded accounts get chopped up. I run every candidate through five checks before I even draw a neckline, and I skip the pattern if any one of them fails.
First, there must be a real trend into the pattern. A triple top without a preceding uptrend is just a sideways ceiling; a triple bottom without a preceding downtrend is just a sideways floor. Second, the three touches must land within a tight band — I allow about 10 pips of spread on GBPUSD H1, and I reject anything wider, because a level that moves 30 pips between taps is not a level at all. Third, the pullbacks between the taps must be meaningful, at least 25 to 30 pips on H1, otherwise the three peaks are one messy cluster rather than three separate attempts. Fourth, volume or momentum should fade on each successive tap: buyers trying less hard each time is exactly what exhaustion looks like. Fifth, the pattern needs time — at least 15 to 20 bars on the trading timeframe, so the three attempts are genuinely separate battles rather than one afternoon of noise.
The check that saves the most money is the tight band. On my missed Wednesday trade, the three peaks were 1.2850, 1.2852 and 1.2848 — a 4-pip band. In my journal, every profitable triple had a band under 10 pips, and every losing one had a band over 20. A sloppy level gives sloppy breaks, and sloppy breaks reverse into your daily limit.
The full rules: setup, entry, stop, targets and session filter
The setup starts on H4, where I confirm the prior trend and mark the level. The trading timeframe is H1, and I never take a triple pattern on M15 or below — the third touch on a low timeframe is usually just one H1 candle chopped into pieces, and the neckline break that follows it is noise dressed as structure.
Entry comes in two approved forms, and I use the retest form by default. The aggressive form enters on the H1 close beyond the neckline: for a triple top, a close below the lowest pullback low; for a triple bottom, a close above the highest relief high. The retest form waits for price to come back to the broken neckline and reject it — a bearish rejection candle at the neckline for a short, a bullish one for a long — and enters on that rejection. The retest entry is slower and sometimes never comes, but it puts my entry 10 to 20 pips closer to my stop, and on a funded account that distance is worth more than the trades I miss waiting for it.
Stops are where most triple-top guides quietly cost you money, so here is the honest version. The conservative stop sits beyond the extreme peak or trough of the pattern, plus a 10-pip buffer: above the highest of the three peaks for a short, below the lowest of the three troughs for a long. That stop is safe, but it is wide — usually wider than the pattern height — which means the full measured move pays less than 1R. The retest-swing stop, which I use, sits beyond the retest high or low instead: above the highest point of the failed retest for a short, below the lowest point of the failed retest for a long, plus the same 10-pip buffer. It is tighter, it pays a proper multiple, and it accepts a real trade-off: price can push past the retest swing and still complete the pattern, stopping me out of a trade that would eventually have worked. I accept that trade-off because a funded account is killed by wide stops, not by missed winners.
Targets follow the classic measured move, split into two parts. Take a partial close at 1R, move the stop to breakeven, and let the remainder run to the measured target — neckline minus pattern height for a short, neckline plus pattern height for a long. If price stalls at the halfway point of the measured move and starts chopping, I take the rest off early rather than watching a winner turn into a breakeven trade. The session filter is simple: trade London and New York hours only, never inside thirty minutes of a red-folder release, and never on a Friday afternoon, when triple patterns love to break one way, retest, and then reverse into the weekend gap.
Position-sizing maths under a 5,000 dollar daily limit
The sizing rule is the same one I use for every strategy on this site, and it is the part that actually protects the account. On a 100,000 dollar funded account with a 5,000 dollar daily loss limit, I risk 500 dollars per trade — 10 percent of the daily limit, so one loss can never end my day and ten consecutive losses still leave the account alive. The formula is risk dollars divided by stop distance in pips times pip value:
| Step | Calculation | Worked value |
|---|---|---|
| Pattern height | Peak minus neckline | 1.2850 minus 1.2780 equals 70 pips |
| Measured target | Neckline minus height (short) | 1.2780 minus 0.0070 equals 1.2710 |
| Stop distance | Entry to stop, retest-swing stop | 1.2778 to 1.2812 equals 34 pips |
| Position size | 500 divided by (34 times 10) | 1.47 lots |
| Loss if stopped | 34 pips times 10 dollars times 1.47 | 500 dollars |
Before you run these numbers on your own account, check them against the max daily drawdown calculator so the trade fits your firm’s actual limit, not the round numbers in this example. And note what happens if you use the conservative peak stop instead: the stop distance jumps to 84 pips, the size falls to 0.60 lots, and the measured target pays only about 0.8R. That is the maths behind my stop choice — it is not a preference, it is the difference between a trade that pays for its risk and one that does not.
| Stop distance | Size at 500 dollars risk | Loss if stopped |
|---|---|---|
| 25 pips | 2.00 lots | 500 dollars |
| 34 pips | 1.47 lots | 500 dollars |
| 50 pips | 1.00 lot | 500 dollars |
| 84 pips (peak stop) | 0.60 lots | 500 dollars |
| 100 pips | 0.50 lots | 500 dollars |
The worked trade: a triple top on GBPUSD
Here is the trade from that Wednesday, reconstructed with the rules above instead of the hesitation I actually traded with. GBPUSD on H1, after a two-week uptrend, tapped 1.2850, pulled back to 1.2780, tapped 1.2852, pulled back to 1.2785, and tapped 1.2848. The band was 4 pips, the pullbacks were 70 and 67 pips, and momentum faded on each tap — all five checks passed. The neckline was the lower pullback low, 1.2780, and price closed below it at 1.2772.
| Trade element | Value |
|---|---|
| Peaks | 1.2850, 1.2852, 1.2848 (4-pip band) |
| Neckline | 1.2780 (lower pullback low) |
| Pattern height | 70 pips |
| Retest entry (short) | 1.2778, on the bearish rejection at the neckline |
| Stop (retest high 1.2802 plus buffer) | 1.2812 (34 pips) |
| Position size: 500 divided by (34 times 10) | 1.47 lots |
| First target TP1 (1R) | 1.2744, plus 500 dollars |
| Second target TP2 (measured move) | 1.2710, plus 1,000 dollars |
Price hit TP1 the same session, I took half off and moved the stop to breakeven, and the remainder reached the measured target the next morning for the full 2R. Every number was decided before entry: the neckline, the stop, the size, both targets, and the partial rule. A triple pattern without those pre-decided numbers is just a pretty chart.
I tested it: ten filtered triple trades against the unfiltered version
Over eight weeks I logged every triple top and triple bottom on GBPUSD and EURUSD H1 that passed all five checks, then split the log the same way I split every strategy test on this site: trades taken with the full rule set — retest entry, retest-swing stop, London or New York hours, no news within thirty minutes — against trades taken on the raw neckline break alone. The filtered log is the only one I would trade with real capital, and the gap between the two logs is the whole argument for the rules:
| Trade | Pattern | Result (R) | Filtered? |
|---|---|---|---|
| 1 | Triple top, GBPUSD | +2R | Yes |
| 2 | Triple bottom, EURUSD | +1R | Yes |
| 3 | Triple top, EURUSD | -1R | Yes |
| 4 | Triple bottom, GBPUSD | +2R | Yes |
| 5 | Triple top, GBPUSD | -1R | Yes |
| 6 | Triple bottom, EURUSD | +1R | Yes |
| 7 | Triple top, GBPUSD | +2R | Yes |
| 8 | Triple bottom, GBPUSD | -1R | Yes |
| 9 | Triple top, EURUSD | +1R | Yes |
| 10 | Triple bottom, GBPUSD | +2R | Yes |
Filtered total: 7 wins, 3 losses, winners 11R against losers 3R, net plus 8R equals plus 4,000 dollars at 500 dollars risk per trade. The best single trade made 1,000 dollars, 25 percent of the net — inside the 30 percent consistency cap most firms apply. The unfiltered log was 10 trades, 4 wins, 6 losses, net minus 2R equals minus 1,000 dollars, mostly from breaks that never retested and patterns with a 30-pip band. Same pattern, same market, opposite outcome — the five checks and the retest entry were the strategy.
When it fails, and how it fits prop firm rules
Triple patterns fail in three predictable ways, and knowing them in advance is what keeps a failure to one 500-dollar loss instead of a breached daily limit. The first failure is the fake break: price closes beyond the neckline, you enter, and price snaps straight back inside the pattern. The retest entry removes most of these, because a fake break rarely produces a clean rejection at the neckline — it usually just drifts back through it. The second failure is the rectangle in disguise: price breaks the neckline, runs a few pips, and then ranges again, because the “pattern” was really a consolidation and the break was just the range expanding. The tight-band and meaningful-pullback checks filter most of these before entry. The third failure is the trend resuming through the third peak: buyers absorb three rounds of selling and break the ceiling anyway, which is why the conservative peak stop exists as a concept even though I do not use it — if you ever trade this pattern without a stop, that third-peak break is the move that ends your account.
Against firm rules, the pattern fits well when traded the filtered way. The daily loss limit is protected by the fixed 500-dollar risk and by the pattern’s rarity — one or two genuine candidates a week, so there is no pressure to overtrade. The consistency rule is protected by the partial-at-1R rule, which stops any single trade dominating a payout period, as the test log shows. The trailing drawdown rule is the one to watch: a runner that reverses can drag a trailing floor up and then breach it, so I move the stop to breakeven at TP1 without exception. If you want the level-flip version of the same idea, the break and retest strategy covers it, and the rectangle breakout strategy is the right comparison when you suspect your triple is really a box.
Frequently asked questions
Is a triple top stronger than a double top?
Yes, but only after the neckline breaks. Three failed attempts mean more trapped traders than two, so a genuine triple break often moves faster. Before the break, a triple is weaker evidence than it looks — the third tap could still just be a range, which is why the neckline close is the trigger, not the third touch itself.
How close together must the three peaks be?
On GBPUSD H1 I allow about 10 pips. Wider than that and the level is not really a level; the breaks that follow it tend to be sloppy, retest poorly, and reverse. On a less volatile pair you can tighten the band further, but never widen it to force a pattern to qualify.
Should I enter on the break or wait for the retest?
Wait for the retest, unless the break is so strong — a full-bodied H1 close well beyond the neckline with momentum — that waiting would mean chasing. The retest entry puts you closer to your stop, which is worth more on a funded account than the occasional runaway break you miss.
Where does the stop go if the retest pushes past the old peak?
Then the trade is invalidated under the retest-swing stop, and you are out for a small loss. The pattern itself may still complete later, but a retest that reaches the old peak has told you the level is not holding as resistance, and trading on after that signal is guessing, not following rules.
Does this pattern work on lower timeframes?
It appears on lower timeframes constantly, and most of those appearances are noise. I trade it on H1 with H4 context only. On M15, three taps are often a single H1 structure chopped into pieces, and the neckline breaks fail at a rate that no sizing rule can fix.
How often should I expect a genuine triple pattern?
On two major pairs, roughly one or two genuine candidates a week that pass all five checks. If you are seeing one every day, your checks are too loose — most likely the band is too wide or the prior trend is missing, and you are trading rectangles.
Risk disclaimer: Trading leveraged products carries a high level of risk and may not be suitable for all investors. The examples in this article use a hypothetical 100,000 dollar funded account and are for education only; they are not trading advice, and past results from any strategy do not guarantee future results. Always check your prop firm’s current rules before trading a new setup.
Further reading on the pattern’s classic form: Triple Tops and Triple Bottoms Explained




